The North American Carbon Budget

2.5.1 Canada, the United States, and Mexico

Efforts to understand the North American carbon cycle—including its stock and flux changes and socioecological implications—cross sociopolitical and economic boundaries. This report shows that regional efforts have measured, modeled, and scaled carbon sources and sinks across North America and quantified the uncertainties associated with those estimates. Arguably, the most carbon cycle information is available for the United States, followed by Canada and Mexico. This information availability translates into higher confidence for estimates of carbon dynamics across the conterminous United States and Canada but lower confidence for Mexican estimates.

In general, SOCCR1 and subsequent publications (see sections above) suggest that terrestrial ecosystems in Mexico act as net sources of carbon to the atmosphere (due to land use and agricultural practices), while those in the United States and Canada tend to be net sinks of carbon from the atmosphere. In contrast, the United States is the highest emitter of fossil fuel emissions, followed by Canada and Mexico. These dynamics are related to differences in socioecological drivers that regulate carbon dynamics among the three countries, influencing the ­continental-scale carbon cycle.

The United States is characterized by a stable forestland, whose area gains and losses have roughly balanced over the last century (see Ch. 9: Forests), enhancing the terrestrial carbon sink. In contrast, the large U.S. economy and population have high energy demands that contribute to the largest carbon emissions in North America. U.S. fossil fuel emissions were 1.5 Pg C per year (±6%) from 2004 to 2013 (see Table 2.2), or approximately 4,700 kilograms (kg) C per person. Canada is characterized by an extensive natural resource base, where forests represent the largest ecosystem carbon pool. These forests have high disturbance rates and low productivity, resulting in an overall nearly neutral carbon balance. Although Canada’s per capita emissions rate of 4,100 kg C is similar to the U.S. rate, its lower population resulted in substantially smaller fossil fuel emissions (148 Tg C per year ± 2%) from 2004 to 2013. In contrast, Mexico is characterized by higher-productivity forests (particularly its tropical forests), but also by more frequent natural disturbances (e.g., droughts, hurricanes, and fires) and high pressure on the use of natural resources that drives land-use change. Mexico contributed 130 Tg C per year (±15%) in fossil fuel emissions from 2004 to 2013, and its per capita emissions rate (1,000 kg C) is much lower than that of the United States and Canada because of its relatively large population with lower energy consumption.

Fully understanding differences in carbon dynamics across North America requires identifying the size of its carbon pools and the influence of climate feedbacks (e.g., changes in temperature or precipitation patterns) on the capacity of the pools to sequester or release carbon. In addition, differences in population migration patterns (e.g., changes between rural and urban populations), along with economic energy demands, determine anthropogenic drivers and feedback mechanisms of carbon exchange across the three countries of North America.

2.5.2 National Climate Assessment Regions of the United States

Terrestrial ecosystems in the eastern United States—located roughly within the Northeast, Midwest, Southeast, and Caribbean National Climate Assessment regions—together have acted as a substantial carbon sink in recent decades (Xiao et al., 2014; Zhu and Reed 2014), largely because of carbon accumulation in forests recovering from past disturbances (Williams et al., 2012). Most of the carbon sink in the eastern United States is in the Northeast and Southeast regions; the carbon sink in the Midwest region is relatively small in comparison. This regional difference is influenced mainly by the dominance of forests in the Northeast and Southeast regions and of agricultural lands in the Midwest. Projected carbon uptake in the Northeast and Southeast regions between 2006 and 2050 is expected to decrease from the current level, primarily because of forest aging in these regions (Liu et al., 2014). A better understanding of forest carbon dynamics is needed to quantify the impacts of 1) forest management, including the locations and intensity of widespread partial cutting in the Northeast region (Zhou et al., 2013); 2) disturbances such as windstorms (Dahal et al., 2014); 3) climate and atmospheric changes including CO₂ fertilization (Norby and Zak 2011); and 4) wildland fires (Turetsky et al., 2014). Forest land uses including harvesting (i.e., clear-cutting and partial cutting, with forests remaining as forests) and conversion to other land uses are important driving forces of carbon cycling, not only for direct immediate carbon removal from these activities, but also for subsequent activity-dependent paths of changes in carbon storage. Although wildland fires have contributed only a small source effect on the total U.S. net carbon balance in recent decades (Chen et al., 2017), the area burned by wildland fires and the associated GHG emissions are projected to increase in the future (Hawbaker and Zhu 2014). Carbon stored in the Atlantic coastal wetlands is particularly vulnerable to wildland fires because of land-use activities (Flores et al., 2011).

Terrestrial ecosystems in the Great Plains region acted as a carbon sink from 2001 to 2005 (Zhu et al., 2011). Their current rate of uptake is expected to remain steady or decrease slightly until 2050 as a result of climate change and projected increases in land use. Methane emissions from wetlands and N₂O emissions from agricultural lands are high for the region and expected to increase. The amount of area burned in the Great Plains and the region’s GHG emissions are highly variable, both spatially and temporally. Although estimates for the amount of area burned are not expected to increase substantially over time, fire-resultant GHG emissions are expected to increase slightly for a range of climate projections. Land-use and land-cover changes are major drivers of shifts in the region’s carbon storage. Consequently, future carbon storage in the Great Plains region will be driven largely by the demand for agricultural commodities, including biofuels, which might result in substantial expansion of agricultural land at the expense of grasslands, shrublands, and forests. Converting these areas to agricultural lands, among other land-use changes, may lead to considerable loss of carbon stocks from Great Plains ecosystems. Moreover, studies have not fully examined the important regional effects of climate variability and change, such as droughts, floods, and fluctuations in temperature and moisture availability.

The western United States, consisting roughly of the Northwest and Southwest climate regions, acted as a net terrestrial carbon sink from 2001 to 2005 (Zhu and Reed 2012). The carbon density in these regions demonstrated high spatial variability in relation to variation along a climate gradient from the Marine West Coast to Warm Desert ecoregions. Furthermore, drought is recognizably important in the interannual variability of carbon dynamics in water-limited ecosystems across the southwestern United States (Schwalm et al., 2012; Biederment et al., 2016). Compared to the region’s contemporary rate of uptake, future carbon sinks in the western United States are projected to decline, mainly in ecosystems of the Northwest region in response to future climate warming and associated drought effects (Liu et al., 2012). Influenced by both climate and land-use changes, wildland fires have been major ecosystem disturbances in the Northwest and Southwest regions (Hawbaker and Zhu 2012), resulting in considerable interannual and regional variability in GHG emissions, mostly in the semiarid and arid Western Cordillera and Cold Desert ecoregions. From 2001 to 2005, average annual GHG emissions from the fires equaled 11.6% of the estimated average rate of carbon uptake by terrestrial ecosystems in the western United States. Under future climates scenarios, areas burned by wildland fires and the associated GHG emissions are projected to increase substantially from the levels of 2001 to 2005. Other ecosystem disturbances, such as climate- and insect-caused forest mortalities, are important drivers of carbon cycling in these regions, but incorporating these processes into regional carbon cycle assessments remains a major challenge (Adams et al., 2013; Anderegg et al., 2013; Hartmann et al., 2015).

Although forestlands of southeastern Alaska are included in national GHG reports, other regions of Alaska are not because field data for them is insufficient to support a formal inventory program and many areas are classified as “unmanaged” according to the Intergovernmental Panel on Climate Change. However, Alaska’s high-latitude ecosystems are potentially more vulnerable to future climate change than regions in the temperate zone because increasing temperatures may expose the substantial stores of carbon in the region to loss from increasing wildfire and permafrost thaw. To better understand these potential effects, researchers conducted a more comprehensive assessment of carbon stocks and fluxes of CO₂ and CH₄ across all ecosystems in Alaska by combining field observations and modeling (McGuire et al., 2016). The assessment found that temperate forests in southeastern Alaska store approximately 1,600 Tg C across the major pools, with about twice as much in live and dead tree biomass (1,000 Tg C) than in the SOC pool (540 Tg C). In contrast, the vast majority of carbon stocks in Alaska’s northern boreal forest and Arctic tundra ecosystems occur in SOC (31 to 72 Pg C), much of which is stored in frozen ground (see Ch. 11: Arctic and Boreal Carbon). Despite the average annual source of 5.1 Tg C from the boreal region due to wildfire, Alaskan upland ecosystems overall were estimated to be, on average, a net sink of 5 Tg C per year over recent decades (1950 to 2009). During the same period, this sink was offset partially by the state’s wetland ecosystems that acted as a net source of 1.3 Tg C per year, including 0.93 Tg C per year in biogenic CH₄ emissions since 2000. Finally, the total net flux from inland waters across Alaska is estimated at approximately 41.2 ± 20 Tg C per year, where total net flux equals coastal export plus CO₂ emissions from rivers and lakes minus burial in lake sediments. However, projections from the Alaska assessment indicate that increased uptake in upland and wetland ecosystems over this century will more than compensate for sources resulting from wildfire, permafrost thaw, and wetland emissions. Carbon sinks in Alaska’s upland and wetland ecosystems are projected to increase substantially (18.2 to 34.4 Tg C per year) from 2010 to 2099, primarily because of a 12% to 30% increase in net primary production associated with responses to rising atmospheric CO₂, increased nitrogen cycling, and longer growing seasons.